(left) Total UMI counts for each spot overlayed on the tissue image. Spots with greater UMI counts likely have higher RNA content than spots with fewer UMI counts.
(right) Total UMI counts for spots displayed by a 2-dimensional embedding produced by the t-SNE algorithm. In this space, pairs of spots that are close to each other have more similar gene expression profiles than spots that are distant from each other.

(left) These are the assignments of each spot-barcode to clusters by an automated clustering algorithm. The clustering groups together spots that have similar expression profiles. In this plot, spots are colored according to their cluster assignment and projected on to the tissue image. Only spots under tissue are used in the clustering algorithm.
(right) Spots are colored by clustering assignment and shown in t-SNE space. The axes correspond to the 2-dimensional embedding produced by the t-SNE algorithm. In this space, pairs of spots that are close to each other have more similar gene expression profiles than spots that are distant from each other. The display is limited to a random subset of spots.